برآورد تبخير و تعرق مرجع با استفاده از داده‌هاي سنجش‌ از دور در دشت همدان بهار

Estimation of reference evapotranspiration using remote sensing data in Hamedan-Bahar Plain

با توجه‌ به افزايش توليد محصولات كشاورزي و همچنين وقوع خشكسالي مكرر در بسياري از مناطق جهان، نياز شديد به برآورد دقيق‌تري از ميزان آب مصرفي گياهان و درنتيجه برآورد دقيق تبخير و تعرق مرجع احساس مي‌شود. معادلة پنمن مانتيث فائو براي برآورد تبخير و تعرق مرجع به‌عنوان روشي استاندارد در بسياري از تحقيقات معرفي شده است. از معايب اصلي اين روش به‌صورت نقطه‌اي و دردسترس‌بودن داده‌هاي هواشناسي در مكان‌هاي خاص مي‌باشد. درحالي‌كه با استفاده از داده‌هاي سنجش‌ازدور مي‌توان اين مشكل را برطرف نمود. در اين پژوهش، هدف اصلي تركيب داده‌هاي سنجش‌ازدور با مدل-هاي يادگيري ماشين براي برآورد تبخير و تعرق مرجع است. با استفاده از مدل‌هاي يادگيري ماشين، چالش‌هاي انتخاب بهترين مدل ممكن، متغيرهاي ورودي مدل و دردسترس‌بودن داده‌هاي موردنياز ايجاد مي‌شود؛ بنابراين در اين پژوهش مدل-هاي مطرح RF، GBR و SVR انتخاب و از داده‌هاي تصاوير لندست و شاخص‌هاي پوشش گياهي استفاده شد. منطقه موردمطالعه، دشت همدان بهار واقع در مناطق غربي كشور است. در اين پژوهش براي برآورد تبخير و تعرق مرجع، از دو رويكرد استفاده شد كه در رويكرد اول، متغيرهاي ورودي مدل‌ها با مقادير همة باندهاي تصاوير لندست، درحالي‌كه در رويكرد دوم، شاخص‌هاي پوشش گياهي به‌عنوان ورودي مدل معرفي و استفاده شد. مدل RF با شاخص‌هاي پوشش گياهي، نتايج آماري برابر با (%14.1=RMSE و %76.4=R2) داشت، درحالي‌كه با استفاده از همة باندهاي لندست (%11.7=RMSE و %84.1=R2) و همچنين در مقايسه با الگوريتم‌هاي ديگر، با مقداري دقت بيشتر تبخير و تعرق مرجع را برآورد كرد. نتايج، بيان‌كننده توانايي و پتانسيل شاخص‌هاي پوشش گياهي به‌تنهايي و تصاوير لندست در تهيه اطلاعات لازم براي مديريت آبياري در كشاورزي و همچنين توانايي الگوريتم‌هاي يادگيري ماشين در برآورد پارامترهايي نظير تبخير و تعرق مرجع مي باشد.

Reference evapotranspiration (ETo) is a major research area of both hydrology and water resources management. The most important and direct application of ETo is in the field of irrigation. One of the conventional methods for estimating reference evapotranspiration using meteorological data is the Penman-Monteith-FAO equation. This equation due to satisfactory results has been used in a variety of climates around the world. However, the lack of necessary meteorological data makes it difficult to estimate spatially distributed ETo using the FAO-PM method in the wider ungauged areas. Penman Monteith method requires the data of air temperature, wind speed, relative humidity, solar radiation and etc. To overcome the existing limits of the FAO-PM model, various attempts aiming to estimate ETo with limited observed data have been conducted. Remote sensing methods are already the only way to obtain the various variables at the temporal and spatial scales that needed to estimate evapotranspiration. In recent years, several algorithms have been proposed to estimate reference evapotranspiration using remote sensing data. Some of these models, which are based on the relationship of energy balance, are called surface energy balance methods. In addition to remote sensing, data analysis techniques based on machine learning (ML) are more frequently used in agricultural studies in recent years, especially in evapotranspiration. Therefore, analyses performed with ML algorithms, when coupled with remote sensing data, have the potential to predict the biophysical variables, mainly due to the adaptive capacity of the models to find patterns in nonlinear behavior variable, such as ETo. Machine learning methods are well known and have been widely used in other engineering sciences. The purpose of this study is to estimate the reference evapotranspiration using machine learning algorithms and remote sensing data, and finally to analyze the algorithms used. In general, the final results of evapotranspiration estimation depend on factors such as the type of data and the method for estimating evapotranspiration. In this study, the standard method of estimating ETo with meteorological data, Penman-Monteith FAO equation was used. The NDVI vegetation index indicates the amount of vegetation on the ground and is sensitive to the early stages of phenology. But the enhanced vegetation index (EVI) minimizes atmospheric effects and differences in blue and red reflections. The SAVI index is used to calculate the vegetation of the land surface that has moderated the effect of soil on it. Three machine learning algorithms were introduced to train the ET0 models, including random forest (RF), gradient boosting regressor (GBR) and support vector regression (SVR). Random forest is one of the machine learning methods that performs classification and regression using Bootstrap and Bagging methods. In this research, three machine learning algorithms with different input data (vegetation indices and all bands of Landsat 7 and 8) were used and after comparing the results, the best model was selected. Performance Evaluation Indicators considered to compare and evaluate the performance of the studied models were the parameters of mean square error (RMSE), mean absolute error (MAE), coefficient of determination (R2) and correlation coefficient (CC). Finally, according to the results of the two approaches used in this study, using the values of all Landsat bands, the reference evapotranspiration can be estimated with more accuracy. Accurate estimating of reference evapotranspiration is necessary to estimate irrigation needs and in general, to accurately manage water resources. Conventional methods of measuring evapotranspiration are reference using meteorological data. These measurements are point-based, so they are only suitable for very small scale areas. At present, remote sensing methods are the only non-terrestrial way to obtain the various variables at the temporal and spatial scales needed to estimate reference evapotranspiration. In order to reduce the dependence on climatic data and better resolution, machine learning methods are used to calculate the reference evapotranspiration. In this research, RF, GBR and SVR models were used. In the present study, two approaches were used. In the first approach, the values ​​of all bands of Landsat images were as model input; while in the second approach, vegetation indices were calculated with only a few bands of Landsat images and then used as model inputs. By examining, it could be seen that the information obtained from the Landsat image bands is related to the phenological behavior of the products, and it is also possible to contract very relevant information related to agricultural products that are examined temporarily and spatially. One of the factors influencing the accuracy of estimating reference evapotranspiration is the use of other Landsat bands in addition to the bands related to vegetation indices.